Objectives are to validate a simple classification for irradiated specimens and assessing the incidence and the outcome of sterilized forms. Between 1996 and 2005, 56 non metastatics patients had preoperative chemoradiation and curative resection for pancreatic adenocarcinoma. We retrospectively applied the Dworak regression scale previously describe for rectal cancer. Dworak 4 (sterilized tumor), 3, 2, 1 and 0 grades interested 7 (12,5%), 12, 12, 11 and 14 patients respectively. The median estimated overall survival of all patients was 24 months with estimated 1-, 3- and 5-year survivals of 80%, 35% and 18% respectively. Statistical analysis permitted to regroup patients classified Dworak 4 or 3 (grade 2 of our modified Dworak classification (MDC)) and Dworak 2, 1 or 0 (grade 1 of our MDC). Patients with grade 2 MDC had an estimated median survival and 5-years survival of 40 months and 28 % respectively. Eleven patients (58%) with grade 2 MDC (n = 19) had exclusive metastatic recurrences. Nineteen patients with grade 1 MDC (n = 37) had metastatic (n = 17 ; 46% ; p = 0,07) or local recurrences (n = 2). The MDC was useful because a) easy to used and b) correlated with good prognostic factor for patients with grade 2 MDC. However, metastatic recurrence rate didn't differed in the 2 groups. Thus, adenocarcinoma of the pancreas had to be treated by surgical curative resection associated with radiotherapy and systemic chemotherapy to control the both side, metastatic and local, of the disease. The best preoperative treatment had to be define but must include CRT and systemic chemotherapy.